To improve realism of targetry training, DSTA explored the use of Augmented Reality (AR) technologies.
Previously, realistic training for forward observers, ground forward air controllers, anti-armour operators and anti-aircraft operators was mainly carried out through field training involving platforms and live-firing, but opportunities to conduct such training were limited.
To provide realistic and engaging training, the team experimented with mock-up sensor and weapon systems with AR capabilities to substitute platforms and live-firings.
Working closely with the Army, the team accomplished several challenging AR capabilities in sub-urban terrains during the trials, which facilitated a quick flow of feedback for the systems to be improved iteratively. For instance, as existing commercial off-the-shelf AR software development kits are unable to perform long-distance estimations accurately, the team developed and fused several algorithms to gain accurate reference points in the terrain. This allowed virtual entities to be correctly registered onto physical spaces.
Additionally, the team successfully tested the occlusion of virtual entities, so that virtual entities positioned behind physical structures (e.g. buildings) are accurately concealed instead of appearing in a soldier’s field of view. The experimentation also displayed the ability of the sensor and weapon systems to interoperate with existing simulation systems, which would allow cross simulation training in the future.